Update README.md
Browse files
README.md
CHANGED
@@ -15,14 +15,23 @@ Replete-LLM-Qwen2-7b
|
|
15 |
|
16 |
Thank you to TensorDock for sponsoring **Replete-LLM**
|
17 |
you can check out their website for cloud compute rental below.
|
18 |
-
-
|
19 |
-
https://tensordock.com
|
20 |
_____________________________________________________________
|
21 |
**Replete-LLM** is **Replete-AI**'s flagship model. We take pride in releasing a fully open-source, low parameter, and competitive AI model that not only surpasses its predecessor **Qwen2-7B-Instruct** in performance, but also competes with (if not surpasses) other flagship models from closed source like **gpt-3.5-turbo**, but also open source models such as **gemma-2-9b-it**
|
22 |
and **Meta-Llama-3.1-8B-Instruct** in terms of overall performance across all fields and categories. You can find the dataset that this model was trained on linked bellow:
|
23 |
|
24 |
- https://huggingface.co/datasets/Replete-AI/Everything_Instruct_8k_context_filtered
|
25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
Some statistics about the data the model was trained on can be found in the image and details bellow, while a more comprehensive look can be found in the model card for the dataset. (linked above):
|
27 |
|
28 |
![image/png](https://cdn-uploads.huggingface.co/production/uploads/642cc1c253e76b4c2286c58e/75SR21J3-zbTGKYbeoBzX.png)
|
|
|
15 |
|
16 |
Thank you to TensorDock for sponsoring **Replete-LLM**
|
17 |
you can check out their website for cloud compute rental below.
|
18 |
+
- https://tensordock.com
|
|
|
19 |
_____________________________________________________________
|
20 |
**Replete-LLM** is **Replete-AI**'s flagship model. We take pride in releasing a fully open-source, low parameter, and competitive AI model that not only surpasses its predecessor **Qwen2-7B-Instruct** in performance, but also competes with (if not surpasses) other flagship models from closed source like **gpt-3.5-turbo**, but also open source models such as **gemma-2-9b-it**
|
21 |
and **Meta-Llama-3.1-8B-Instruct** in terms of overall performance across all fields and categories. You can find the dataset that this model was trained on linked bellow:
|
22 |
|
23 |
- https://huggingface.co/datasets/Replete-AI/Everything_Instruct_8k_context_filtered
|
24 |
|
25 |
+
Try bartowski's quantizations:
|
26 |
+
|
27 |
+
- https://huggingface.co/bartowski/Replete-LLM-Qwen2-7b-exl2
|
28 |
+
|
29 |
+
- https://huggingface.co/bartowski/Replete-LLM-Qwen2-7b-GGUF
|
30 |
+
|
31 |
+
Cant run the model locally? Well then use the huggingface space instead:
|
32 |
+
|
33 |
+
- https://huggingface.co/spaces/rombodawg/Replete-LLM-Qwen2-7b
|
34 |
+
|
35 |
Some statistics about the data the model was trained on can be found in the image and details bellow, while a more comprehensive look can be found in the model card for the dataset. (linked above):
|
36 |
|
37 |
![image/png](https://cdn-uploads.huggingface.co/production/uploads/642cc1c253e76b4c2286c58e/75SR21J3-zbTGKYbeoBzX.png)
|